
Founder & CEO, Sonarium Labs
Telecom ParisTech Graduate
Previously at




10+ years building production ML systems for the kind of problems energy platforms run on: forecasting under uncertainty, optimization with real constraints, and decision support where being wrong costs money.
At BNP Paribas, quantitative finance models — the same mathematical toolkit behind energy trading, hedging, and PPA valuation. At Rappi, logistics optimization at scale across Latin America — structurally similar to grid balancing and asset optimization.
He founded Sonarium Labs to apply that experience to energy — building and evaluating AI systems where better models translate directly into measurable outcomes.
“The questions that matter aren't usually in the pitch deck. Is the forecasting accuracy real? Is the optimization solving the stated problem? What breaks when the team that built it leaves?”— Nicolas Debaene, Founder of Sonarium Labs
We build and evaluate AI systems for energy — from forecasting models to production pipelines — across three dimensions that matter.
Is the model performance real — or measured against a weak baseline?
Why It Matters
Energy platforms where a model that's 95% accurate on average can be worthless during the 5% of hours that drive P&L.
What We Evaluate
The best ML systems aren't the most sophisticated — they're the ones operators trust and use.
Why It Matters
Regulated energy markets where decisions are auditable, and explainability isn't a nice-to-have.
What We Evaluate
Full pipeline assessment: provenance, access controls, lineage, and regulatory posture.
Why It Matters
Critical infrastructure platforms handling sensitive operational data with GDPR, NIS2, and sector-specific compliance requirements.
What We Evaluate
Tell us about your project and we'll get back to you soon.
Your information is secure and will only be used to respond to your inquiry.